El Hierro – now up to 41 hours of 100% renewables generation

Or to be exact, 40 hours and 40 minutes of 100% renewables generation between 1 am on February 14 and 17.40 on February 15. But still enough to eclipse the 33 hours of 100% renewables generation achieved at King Island, Tasmania, last November. So congratulations to the GdV project staff.

Here are the REE ten-minute generation values for the EL Hierro grid over the period of interest:

Except for a small amount of hydro generation when wind generation dropped off in the early afternoon of February 14 El Hierro’s electricity demand was filled entirely by wind. (Wind speeds at El Hierro airport 3 km to the northeast ranged from 8 to 13 m/s over the period). This is in contrast to the 16 hours of 100% renewables generation achieved on January 16, when the wind abruptly died and the test continued with a mixture of wind and hydro (see Figure below). Otherwise the only operational change since the January seems to have been that wind was curtailed at around 7MW rather than around 6MW:

My preliminary questions are (others may occur to me later):

1. Why was the test terminated at 17.40 on February 15? Except for a tiny amount of hydro generation when the wind briefly fell off at 17.10 (barely visible on the first plot) there was no significant change in the generation mix, and 41 hours and 40 minutes is hardly a nice round number.

2. Grid stability during the test was probably maintained by using pumping as a dynamic resistor to soak up excess wind and by spinning three of the 2.83 MW hydro turbines to achieve the necessary system inertia. But if it’s that simple, why don’t they do it all the time?

This entry was posted in Energy and tagged , , , , . Bookmark the permalink.

116 Responses to El Hierro – now up to 41 hours of 100% renewables generation

  1. Peter Lang says:


    I’d be interested in updates on the average capacity factor since commercial operations began.

    • sod says:

      “I’d be interested in updates on the average capacity factor since commercial operations began.”

      I have serious doubts about capacity factor as a concept in general. But in this case, with people doing only experiments and wind curtailed at slightly above half of maximum output at all time, the results will be basically meaningless (unless you want to compare to nuclear capacity factor in Japan over the last 2 years..).

      Even looking at the capacity factors of the different months will make little sense, as we wont know which changes are caused by weather, which by choice (and they are making weird choices) and which by a new possibility.
      The best information capacity could give, would be an assessment of water actually having a negative value over prolonged periods of time (the water is obviously simply vanishing from the electricity production cycle).

      I will try to answer both questions from my point of view:

      “1. Why was the test terminated at 17.40 on February 15? Except for a tiny amount of hydro generation when the wind briefly fell off at 17.10 (barely visible on the first plot) there was no significant change in the generation mix, and 41 hours and 40 minutes is hardly a nice round number.”

      Obviously their target was to get above a full day and possibly to beat King island.

      I also assume that these tests require a lot of people being present and most likely a rough target termination point was agreed for the experiment (early evening hours, by the way. Looks like they just called it a day)

      “2. Grid stability during the test was probably maintained by using pumping as a dynamic resistor to soak up excess wind and by spinning three of the 2.83 MW hydro turbines to achieve the necessary system inertia. But if it’s that simple, why don’t they do it all the time?”

      As Rainer explained on the last topic, a slow and careful approach seems to be the mentality of the island and the people running the project.


      I would assume that they only dare to do this, when they have a lot of staff on site and under controlled conditions so far. The second plot above (the 16 h test) obviously shows a switch to hydro at 12 o clock, most likely probing grid stability by such a extreme change (i doubt that such sudden wind changes are even possible).

      I also have a couple of questions/wishes:

      1. How much wind is wasted, if we curtail the system to 60% (7MW)?

      2. What percentage of time do we see wind curtailed at 7MW (or 6 MW before)?

      3. what is the ratio of water being pumped up to wards the number of going down for the different months?

      4. would it be possible to extend the first graph towards the past and present, so that we can see a full week?

      • Roberto says:

        ”unless you want to compare to nuclear capacity factor in Japan over the last 2 years..)”

        I know how much that is!… it is equal to the CF for PV during each and every night of the year, at least 12 hours/day on average.

        See?… That was easy! 🙂

    • Peter: Can do, but I’ll be gone most of the rest of the day so it won’t be until later.

      • willem post says:

        Why look at tea leaves, etc.

        Could we just call them and get a first hand report of what was done and why?

    • Peter: Here are wind capacity factors based on my best estimates of wind power delivered to the El Hierro grid from the 11.5MW El Hierro wind farm from project startup in late June through the end of 2015. They don’t allow for wind curtailments, which were substantial during July and August, and they could include a small amount of hydro that we can’t break out because the hydro numbers provided by REE are the net of uphill pumping and hydro generation and don’t show any hydro generation when the total is negative, as it almost always is. So while they will be in the ball park they won’t be exact:

      June/July …… 23.3%
      August ………. 22.6%
      September ……. 8.7%
      October ……….. 5.2%
      November …… 10.2%
      December ……. 7.5%
      Year …………. 13.2%

      • Peter Lang says:

        Roger Andrews,

        Thank you. Do I understand correctly that the figures represent the overall capacity factor of the renewable energy system for those months? So, as best as you can calculate it given not all the needed data is available, it is total electricity supplied from the wind farms after losses in the pumped hydro system divided by the nameplate generating capacity of the wind farms?

        • Roger Andrews says:

          Peter: The CFs are calculated based on the 11.5MW of installed wind capacity only. Adding the 11.3MW of hydro capacity, which as far as we can tell seems to have produced very little power since startup, would roughly halve them.

          • Peter Lang says:

            Thank you. I understand now (I think). And from my perspective what you’ve said is the correct way to do it. The wind power is the only renewable energy generator. The hydro is simply storage so it doesn’t generate any new energy. Therefore, the wind farm capacity (x hours) is the denominator. The numerator is the output from the wind farms that was sent to the grid, not to storage, plus the energy generated from the pumped hydro that was supplied from wind farms but not from the diesel generators. I expect it would require some advanced statistics capabilities to estimate the numerator and put confidence limits on it. If you wanted it, Joe Wheatley (Biospherica, Ireland) is one person who could do that if it is not yours or Euan’s strong suit.

  2. Rainer says:

    just past the questions a GDV:
    Gorona del Viento
    Cristina Morales Clavijo
    Departamento de Comunicacion


    • Roger Andrews says:

      Rainer: Thank you, but I already know about Cristina. I was in touch with her a few months ago.

      • Rainer says:

        I am an optimist.
        I gave it a new try.
        It is her f….g job to give answers

        • willem post says:

          I think there may be standing instructions not to answer any questions from nosy people who might find what some of them likely already know, i.e., another not-so-well-thought-out RE boondoggle rah-rah-ed through the “system”.

          Some important insider folks may have gotten juicy contracts.

          Lord knows what press releases they have put out, and how much back paddling they would have to do, if and when the truth surfaces.

  3. Euan Mearns says:

    Its difficult to understand. They’ve been pumping for 41 hours. The top reservoir must be full, even if they are allowing the water to circulate down. They’ve still got 7 MW of constrained wind being produced, but they switch on the diesels.

  4. gweberbv says:

    To me it looks they were testing the response of the system to the abrupt outage/slow down of one of wind turbines. See the three dips every 14 hours.

    • Euan Mearns says:

      The second two of these dips have near identical profiles.

    • Roger Andrews says:

      Guenther: Good catch! That’s undoubtedly what they were doing. It also seems that each test was tougher than the one before. They may even have briefly shut a number of turbines down during the third test (wind generation at 17.10 was down to only 4.8 MW – 0.1MW less than demand). An unsatisfactory response to this test may have been why they turned the diesels back on half an hour later.

  5. Seattle is almost entirely renewable energy 24 hours a day, 365 days a year. It’s a fact that 100% renewable is possible, depending on availability of local resources. But that’s the problem. Few places have Seattle’s hydro and wind resources. Indianapolis uses coal for a reason. These island tests are meaningful only for these islands. The results can’t be extrapolated to places like Indiana or Germany. There are wind turbines on some of the Galapagos islands backed up with diesel. Putting reservoirs on those islands would necessitate the destruction of parts of the preserves. These tests don’t really prove much of anything. Build enough reservoirs, pumps, and turbines, and you can become 100% renewable. If only economic feasibility didn’t matter.

    • These island tests are meaningful only for these islands. The results can’t be extrapolated to places like Indiana or Germany.

      Why not?

    • sod says:

      “Seattle is almost entirely renewable energy 24 hours a day, 365 days a year. It’s a fact that 100% renewable is possible, depending on availability of local resources. But that’s the problem. Few places have Seattle’s hydro and wind resources. ”

      You can not compare Seattle, as the town is part of a bigger grid. What is possible in Seattle, will not work everywhere.

      What is possible on El Hierro, will work at every place with similar conditions. And if you look at the King Island system, you have a system that can work basically everywhere.

      El Hierro has just demonstrated, that wind alone can keep the grid stable. And 40 hours is enough to show, that this could go on as long as there is wind.

      “These island tests are meaningful only for these islands. The results can’t be extrapolated to places like Indiana or Germany.”

      Of course it can be extrapolated. And we should definitely do so! In the future, we will need stable microgrid cells inside the country grids. We need this, as our reliance on electricity is increasing and so is the vulnerability in case of an attack/disaster. In contrast to the claims being made, this experiment has demonstrated taht alternative power can be part of the solution, instead of part of the problem.

      ” These tests don’t really prove much of anything. Build enough reservoirs, pumps, and turbines, and you can become 100% renewable. If only economic feasibility didn’t matter.”

      I am not sure, if you really understand this tests. The most obvious point is, that these wind power is massively curtailed. And water is basically not used to produce electricity at all.
      So the economics in this case are much better than we see today. And obviously economics are not a pressing issue, as they are wasting so much wind power.

      And the interesting thing is, that these places can not only generate reliable electricity from (mostly) 100% renewables, but they can even save money doing so.


      • Owen says:

        You are all focusing on the generation side and forgetting completely about the demand side.

        At one extreme, if you take an island with a couple of mud huts and and a shop that opens for 2 hours every day with no refrigeration, then yes anything is possible. You could go 100% renewables very easily.

        At the other extreme, you take a heavily populated and industrialized society with aluminium factories requiring constant electricity with exact frequency then you have a problem.

        Unless of course you outsource the industry and export the demand somewhere else.

        P.S. there are ways out though e.g. off grid generators which will not be counted in your overall electricity figures

  6. Euan Mearns says:

    I see the wind is dropping now, 5 MW (ignoring pumping) is below the 6 MW anticipated evening peak demand. So perhaps the diesels were started in anticipation of wind declining?

    • Rainer says:

      Or just a demand following test of the Enercons E-70.
      I am sure they can do it.
      “Closed-loop control of the wind farm at the point of connection
      A wind farm controller can be used to accurately control the active and reactive power of a wind farm at the point of connection (PoC). In conjunction with the wide reactive power range of ENERCON WECs, a wind farm controller can be used for rapid control of the voltage at the network connection point. ENERCON offers two products for wind farm regulation, which differ with respect to speed and type of control: ENERCON SCADA RTU-C and ENERCON SCADA FCU. With these, the setpoints for control can not only be programmed specifically into the relevant device but also be received dynamically from the grid operator via a communication interface. This enables optimal integration of the wind farm into existing gridsI am not working for Enercon. But if we try to analyse the system we got to look in the brain of the system

    • Kees van der Pool says:

      Hi Euan,
      Water and not power generation is the overriding priority of GdV so until we know the levels of the upper and lower reservoirs at all times, we can only guess. There are patents pending so we should know all about it, in detail, when these are assigned.

      • Rainer says:

        “There are patents pending”

        • Kees van der Pool says:

          Look it up. Its in one of the press releases of the first ‘100%’ event when dignitaries & visitors were kindly requested not to take pictures of the screens in the control room for ‘proprietary’ reasons.

  7. Lars says:

    “40 hours and 40 minutes” of testing.

    Sounds like they have chosen the biblical number for “tribulations”, maybe that`s what this system is going through 🙂

    40 years in the desert, 40 days in the desert tested by the Devil, a woman is 40 weeks pregnant and so forth…

  8. Owen says:

    Easy to do it on a warm island with no industry and basic requirement for static frequency.

    If there was a blackout, how many would notice / care ?

    • sod says:

      “Easy to do it on a warm island with no industry and basic requirement for static frequency.

      If there was a blackout, how many would notice / care ?”

      Sorry, but that is simply wrong. There are no blackouts. It is also most likely, that in a small grid like this one, single users can have a bigger impact. While a lot of industry in other countries has serious back up.

      And obviously it is not that simple to do. If it was, there would not be these experiments and other islands would have such systems for a long time.

      • Owen says:

        Just did a google search which perhaps backs up my point that El Hierro is in no great need of consistent frequency in its electricity generation. However, it does sound like a nice place to visit :

        Tourism has not, really, reached El Hierro. Among the good things are almost empty (but mostly good) roads, a clean environment and almost total peace. No aircraft routes overfly the island and there is no industry apart from some quarries. But there seems a dearth of restaurants and bars with few offering terraces or outdoor areas. The cuisine is almost exclusively Canarian/Spanish. There is a reasonable supermarket on the southern edge of the town of Valverde but most of the villages just have a small grocery shop which keeps short hours. There is a tourist office in Valverde which we didn’t actually find during our visit, although we looked!

  9. Rainer says:

    Since 10:20 ca 7MW wind power. 5th Converter ready again? Ready for next test?
    I think we should be a little more imaginatively.
    Or just like all ways filling the giant water slide mostly by diesel power? El Hierro Disney Waterworld?New idea de “Oficina de Turismo”?

  10. sod says:

    If we look at the data of today (17:00 of 17/02/2016) again, this could have been another day with 100% renewables (90+% wind) easily.


    It is utterly unclear what they are doing.

    Roger gave some capacity numbers above, which are just plain out insane:

    June/July …… 23.3%
    August ………. 22.6%
    September ……. 8.7%
    October ……….. 5.2%
    November …… 10.2%
    December ……. 7.5%
    Year …………. 13.2%

    This is mostly, because wind is curtailed to (now mostly) 7MW and really just never gets close to the 11.5 MW that the system can provide.

    The real capacity factor should be above 50% and the system should mostly work by putting those wind peaks into water storage.

    We must also remember, that there is more wind in summer than in Winter, Rainer provided this link in another topic:


    I could not find any number about the amount of wind being loss by a curtailment to about 60% (7 of 11.5MW), but i found this spaghetti graph from australia (onshore, so El Hierro should outperform the all of them)

    (you can look at other months by simply changing the web adress, el Hierro should perform like a single one of those plants, not like the black average)

    • Roger Andrews says:


      There’s nothing insane or unclear about any of this. Wind capacity factors since August were low simply because of a lack of wind, and when the wind did blow strongly there was significant curtailment, which is inevitable when you have 11.5MW of wind capacity and only ~5Mw of demand and nowhere to store the surplus.

      If you look at the first figure in the post you will see that roughly a third of the wind power during the no-diesel period was “wasted” even with wind curtailed at ~6MW. If it hadn’t been curtailed at this level wastage would have been even higher (how much higher? Try to imagine what wind generation would have looked like if it hadn’t been curtailed. It could easily have exceeded 50%).

      The reason wind capacity factors are higher in places like the UK is that wind penetration is much lower (and the wind evidently more reliable). As a result there are generally no serious problems admitting all of the wind generation directly to the UK grid, although a small amount of curtailment does occur. But if the UK ever did try to supply all of its demand with wind the level of curtailment would increase drastically and we would see much lower wind capacity factors, just as we do at GdV.

  11. Roger Andrews says:

    Two recent comments on this post have prompted me to present the big picture as I see it:

    First this comment from biodiverdist:

    These island tests are meaningful only for these islands. The results can’t be extrapolated to places like Indiana or Germany.

    To which I responded. “Why not?”

    And sod gave me the answer I was looking for:

    What is possible on El Hierro, will work at every place with similar conditions. And if you look at the King Island system, you have a system that can work basically everywhere.

    Sod is correct, except that conditions everywhere are in fact similar to those on El Hierro, where the existence of the volcanic crater now being used as the upper reservoir doesn’t help much.

    This is why I am spending so much time on the GdV project. It’s the future of intermittent renewable energy in microcosm. What works at GdV (and King Island) will indeed work basically everywhere, and what doesn’t work there won’t work anywhere else either – unless of course there’s lots of backup hydro capacity to handle the swings in wind & solar generation, but in these cases there’s no point installing large amounts of wind and solar capacity in the first place.

    The fact of the matter is that high-penetration renewables systems in places like Germany and Indiana, and inplaces like South Australia, Hawaii and France, will face the same two fundamental problems as El Hierro and King Island:

    1. Maintaining grid stability.

    2. Storing excess wind (or solar) energy for use during periods when the wind doesn’t blow and/or the sun doesn’t shine.

    In the last two posts we’ve been discussing problem 1 – grid stability during periods of high wind generation – basically because we haven’t looked at it before. And I get the impression that the problem, if not already solved, is on its way to solution. King Island achieves the necessary system inertia with a dynamic resistor to soak up excess wind power, plus 1.5MWh of battery energy storage, a diesel-based uninterruptable power supply, flywheels and a smart grid/advanced control system. GdV seems to have adopted a similar approach, with uphill pumping being used as a de-facto dynamic resistor to soak up excess wind and with three of the five Pelton hydro generators kept spinning during periods of zero diesel generation to maintain system inertia, backed up again by smart gadgets. So far so good.

    Problem 2, however, is the big one, and the GdV results to date confirm that barring a divinely-inspired leap forward in battery technology it’s unsolvable. Why? For the simple reason that it’s physically impossible to construct storage systems large enough to store energy during high wind (or solar) period for re-use during “slack” periods no matter where you are, a point we’ve made in a number of previous posts. (To “smooth out” the wind fluctuations to date at GdV would require up to 3GWh of storage, 20 times the capacity of the existing reservoirs.) So fossil-fuel backup will always be needed to fill demand when the wind doesn’t blow and the sun doesn’t shine, and the amount of intermittent RE that can be admitted to the grid will decrease as a result. (Based on GdV results to date Hubert Flocard and I estimate that El Hierro as currently configured will be able to generate only 50-55% of its electricity from wind even if all of the usable wind energy generation can be admitted to the El Hierro grid).

    The other question is cost. I haven’t looked into this in detail, but according to published estimates the renewable energy GdV supplied to the El Hierro grid in 2015 (which filled less than a third of El Hierro’s demand) will be billed at an effective rate of around 0.80 euros/kWh. I have no numbers on King Island but suspect they may be in the same range, which could be why I have yet to receive a response from Hydro Tasmania on my request for data.

    • Peter Lang says:


      This is why I am spending so much time on the GdV project. It’s the future of intermittent renewable energy in microcosm.

      I agree. This work you are doing on El Hierro and King Island Tasmania is enormously enlightening and invaluable. It is a very clear demonstration of why renewables cannot provide much of the world’s electricity. I haven’t yet seen someone calculate how many sites equivalent to El Hierro would be required to power the world with weather-dependent renewables.

      • willem post says:


        You are right about RE not contributing much.

        Using at least 10,000 sq miles in the US southwest, CSP with 10-h storage could provide up to 3000 TWh/y to the US economy, with help of a nationwide HVDC system, at about $8 million per mile. Energy cost would be about 15 c/kWh, plus transmission cost. That would be at least 3 times current wholesale prices.

        CSP in Morocco could inflict the same on Europe.

        There likely are other places in the world.

    • gweberbv says:


      I strongly disagree that one can learn a lot from such tiny island grids. The reason is that for such islands interconnectors are not feasable (if they were, we would not call them island grids). However, having interconnectors is by far the cheapest way to achieve a greater utilization of wind and solar. Thus, when integrating renewables the islands grid are fighting with (at least) one hand bound behind their back.

      If you are sitting on an island with 10 MW of wind capacity and 5 MW of demand it is not a realistic option to build a – let’s say – 2.5 MW interconnector to an island 1000 miles away (roughly the distance Dublin-Berlin) that is in the same situation. But if your island has a demand of 50 GW and a wind fleet of comparable size, then it is doable fo build also 10 or 20 GW of interconnectors.
      Still with a distance of about 1000 miles you have roughly the same seasonal variation and also some weather phenomena might cover this area leading to coincident over-/underproduction. But for most of the time, you will enjoy a significant smoothing of wind/solar generation and also to a lesser extend of demand. Resulting in a higher utilization/penetration of the renewables production.
      This is no solution for 100% renewables, but it might bring you from 50% (which you state as the maximum for El Hierro) to 65% or 70%. Actually, the data presented in this blog on wind production in several European countries prove that. 🙂

  12. Graeme No.3 says:

    your work is very enlightening. I would make 2 comments..
    King Island is not the success that sod implies.
    How would South Australia adopt this scheme, seeing it is the driest state in the driest continent?
    No hydro and little change of getting any, short of building vast ponds on the plateau with the Nullabor and cycling sea water. Enormously expensive and the State is close to bankruptcy, and the cost of renewable energy has driven all but 2 large companies from the State, and they are threatening to shut very soon.

    At least the government has started discussing the establishment of a nuclear waste dump to provide some future financial return. The greens are resisting ferociously, trying to end discussion.

    • Roger Andrews says:

      Graeme: South Australia would have to do it the King Island way. King Island has no hydro either.

      You say King Island is not a success. I’m looking for data. Do you have any?

  13. Kees van der Pool says:

    Uh oh – @ 05:00 something just went ’tilt’, load @ zero, Rainer in the dark.

    Trouble as of 01:00, pumping / hydro oscillating due to fast load excursions of 1 MW and it seems the system threw in the towel.

    Press release?
    Maybe not.

  14. Kees van der Pool says:

    The wind became very unsettled @ 01:00, sharp drop, popped up again and was practically gone at @ 05:40. To add to the mayhem, the lone diesel also quit @ 05:20 but was started up again rapidly.

    • Roger Andrews says:

      Was that a planned outage or a genuine system crash? If the latter it took them four hours to recover from it.

      • Kees van der Pool says:

        System crash. I watched the whole thing unfold @ about 9.40pm Pacific Standard Time. The load went straight down to zero, popped up a bit and then drooped again. The graph does not show the hard zero @ 05:00 any more, its ‘smoothed’ out to a flat at 1 MW. Drat – I should have saved the page.

        • The zero is still there in the generation data:

          The predicted demand line (“real” demand doesn’t mean much at the moment) makes it look like they have been having problems since late last night, and they’re still not out of the woods.The interesting thing is how wind generation collapsed before both crashes. According to Rainer’s data (danke Rainer) this was not related to wind speed changes at the airport.

          Could rain be doing this?

          • Kees van der Pool says:

            Thanks for the graph. I think it may have started with rain/wind damaged lines and attempts to re-connect (1MW deltas), more or less what is happening now after the 10:30 mishap. Wind power became really erratic and the pumping/hydro went wild, trying to regulate between rapidly changing loads and violent wind power excursions. No wonder the lone diesel gave up.

            Yes, its the weather. Power grids that live by the weather (etc). . . . . . . .

          • Kees:

            Why does wind power crash first?

            And why are they sending wind to pumping (negative hydro) when they have a ~2MW demand deficit?

  15. Rainer says:

    I confess, just slept during the interest events……
    Inside experience: Rigth now just the weather, heavy rain storm for blackouts. Nothing really special here.
    But first time this winter.
    9:31 airport wind: from 7 m/s to 13m/s with gusts of 20 m/s.
    Happy not to travel today.

  16. Rainer says:

    Airport wind update every 15min. That is new.

  17. Euan Mearns says:

    By coincidence, managed to grab a shot of the wind at 06:00. A wrinkle in an isobar sending wind into irregular speed and direction.

    I doubt that 100% no electricity will have been on the test agenda. And so the main question, will this count as a blackout or a power cut? Perhaps there was a lightning strike.

    They used their hydro in recovery from the Black Start.

    Rainer, how common are “blackouts” on El Hierro?

    • Rainer says:

      Storm and heavy rain:
      may happen any time
      In normal winter if i remeber correct, it was 2-3 times.
      But I may mix it up because the local line of us last year had problems……
      was repaired in ca 20 h

  18. Euan Mearns says:

    A slightly different perspective. Just look at the size of that high pressure, centred on the Azores with low over S Greenland. This is an expression of the NAO – North Atlantic Oscillation


    And the SSTs show a cool N Atlantic, that may herald the cool phase of the AMO – Atlantic multi-decadal oscillation. Also note the fast receding El Nino.

    • Kees van der Pool says:

      El Niño: yes, wisps of cold upwelling started at the Colombian coast. It is raining (SF Bay Area) this morning but El Niño has been a disappointment in February. The snow levels in the Sierra are above normal, though and I’m hoping for a good March.

  19. Rainer says:

    Hello World,
    second Blackout, at least El Golfo, today

    One wlan Hotspot still down.
    Had to change to other Wlan hotspot.
    Here the Voltage protocol:18.02.16 09:21 214
    09:26 224
    09:36 222
    10:27 last internet comunicacion
    10:36 0
    10:37 0
    10:53 0
    11:08 0
    11:09 214
    11:10 214
    11:11 211
    11:12 209
    11:13 213
    11:14 208
    11:15 210
    11:17 213
    11:23 217
    11:26 216

    • Euan Mearns says:

      Yes, they lost wind completely again 10:40. It looks to be a chaotic mess. Why do they not simply fire up the diesels?

      • Rainer says:

        Los Herrenios would say:
        “the muchachos are making a good job repairing the overland lines”.
        They are right and calm enough living isolated, like all people in that situation.
        They do not wait for a blackout, living with really blackouts without problem.
        Diesel does not help to solve problems of the power lines.
        We only have one line to one area. Formula fail safe n-1 is a problem if n=1.

  20. Euan Mearns says:

    Quite happy to have another post called “El Hierro, 100% no Electricity” 😉 Looks like problems with the wind or the wind turbines

    • Euan Mearns says:

      Its now 18:00 in EH and the wind has got back up to 7 MW but they have also switched on a second diesel. So they are now producing 10.1 MW to satisfy about 5 MW demand. I suspect the underlying problem is the ratio of synchronous to non-synchronous supply on the system combined maybe with rain water getting into the turbines.

      • Kees van der Pool says:

        Into the wind turbines, the things I call windmills, right?

      • Euan: All the info you need to figure it out in a single graph 😉

        • Euan Mearns says:

          Amazing chart Roger! We should start trying to sell these things as ltd editions works of art. For me this a thing of pure beauty! They’ve got control back with most power coming from the diesels, some of the wind being used but most being spilled in the water go round.

          Strong wind for a couple of days and then another big dip coming. But something has changed at GdV. They seem now to be trying to use the assets the way they were supposed to be used. My guess is that they now go back to a very cautious routine.

          They need to build the biggest kiddies water slide on EH!

  21. sod says:

    I also missed it.

    But as far as i see it, this blackout should have other reasons than a small dip in wind.

    If they can not handle small wind dips, they should not use wind turbines at all.

    The system was down for nearly one hour. That is strange, as they have three independent systems that could handle load at that time alone.

    Water power should be the perfect way to keep the grid stable or to bring it back up, but obviously they failed. Either they are extremely incompetent, or there was a technical problem that we still do not know about.

    I think a really telling aspect is, that they were pumping water up, while the grid is down. It is rather obvious that they have a completely wrong priority (pumping water up instead of using water to produce electricity), and this shows here again, even if only by accident.

  22. gweberbv says:

    With “apagon canarias” you receive 100,000 results in Google.

    Just on Tueday 200,000 people on Gran Canaria were cut off from the grid. Probably because someone pluged in a solar cell, I suppose.

    • Rainer says:

      You all should calm down a little,
      Please read my comment of:
      February 18, 2016 at 1:47 pm
      Blackouts with this weather conditions are very common here.

      • Rainer: If you check the REE data you will find that this is the first island-wide apagon since GdV went into full operation in June 2015 and probably the first since the plant was inaugurated in June 2014, although I would have to do some more work to verify that.

  23. guber says:

    Well it looks fromm the graph as if the diesel died. At 5:00 there is still wind which delivered some power, while the diesel already delivered zero. That they did not switch to hydro when the diesel died is most likely a software problem, maybe because operation mode is sill relying primary on diesel, so ther was no allowence to automatically switch on hydro.

  24. Kees van der Pool has sent me his interpretation of the sequence of events leading to the outage(s):

    • Kees van der Pool says:

      Another strange sequence is playing out. It started at about 03:00 on 2/19, again preceded by large variations of greater than 1MW of the load. The wind component looked solid but disappeared completely from 03:10 through 04:20, and then spiked up again. I hope this is a test run otherwise it will be difficult to find a reasonable scenario. Hasta mañana.

    • guber says:

      well, at the second occasion there is no drop in wind speed visible which could explain the switch of of Wind. So it’s most likely a different cause.

      Also the 6m/s at 10m height are to be questioned as cause. At 21:00 there was a similar wind speed, and output was still at 4 MW. Other possibility is a common cause which caused trip of first Diesel and then Wind. A common cause e.g. several sever grid faults would also explain why Hydropower did tot show up to compensate.

      The wind speed at airoprt is measured at 10m above ground, wind speed on the mountain and on the tower should be significant higher http://static.twoday.net/QUH/images/wind-hoehe.jpg at 14m/s windspeed close to ground level the 5 wind power generators surely have a output of 11,5 MW, not 7MW. Why they think they need a primary reserve of 4,5 MW remainunknown.

      • Kees van der Pool says:

        The curtailing beats me, too. Regarding the wind dip at 21:00, there was no pumping going on and the dip was beautifully compensated by hydro. At the ‘event’ there was pumping when the wind component disappeared for whatever reason. At that point, the diesel also gave up and hydro did not kick in. Its all pretty mysterious unless the ‘event’ made all the breakers disconnect everything and the system had to be gradually turned on again.

  25. Euan Mearns says:

    So they shut down the second diesel at 23:20 and lost the wind park again at 03:00. This looks like a cabling problem from the wind park. Got to take note of what Rainer is saying, power cuts are routine with wet windy weather.

    • Rainer says:

      Good Morning Folks,
      overslept the 3:00 event also…
      But checked the 2012 Graph, the other years amy follow:,
      2012.08.28 13:20-15:40
      2012.09.24 14:40-20:30
      2012.09.24 00:50-01:10
      2012.11.02 13:50-15:20
      This things are really common.
      Sure more ofthem more local i did not see.

    • power cuts are routine with wet windy weather.

      As they are here in Mexico and in many other places.

      But this one isn’t like previous power cuts. It’s the first outage on El Hierro since GdV went into full operation. All they had to do to recover from previous outages was bring the diesels back up again. Now they have to perform a complicated juggling act between diesels, wind and hydro, and so far they haven’t been doing too well. The grid continues to show signs of periodic instability

  26. sod says:

    Even IF wind power (not the storm!) was the cause (which is utterly unclear, as the wind speed data simply does not show anything really unusual!), it will have few consequences.

    Because as i argued all the time, the system might need minor adjustments only (look at the King Island system). Basically it might need a little bit of battery (as the hydro reaction time might be too slow) and the spin wheels to keep the diesel ready and the grid frequency stable.


  27. Kees van der Pool says:

    The hydro reaction time (Pelton generators) should be pretty much immediate as they are already spinning, either ‘dry’ or ‘wet’ and their 11.3 MW are available within seconds. This is El Hierro’s unique ‘thing’ which stabilizes the system with spinning inertia and is written up here:

    “3.2. No-flow operation of Pelton turbine-driven generators
    In this paper, a new method is proposed to reduce the time
    needed to generate power in Pelton turbine-driven generators.
    This new method proposes synchronizing the Pelton turbine-driven
    generators with the power system as is normally done (Fig. 1.
    Start). First, switch on the turbine controller, which will open the
    injector so that the unit will accelerate to the rated speed. Next,
    switch on the automatic voltage regulator, so that the generator
    voltage will rise to its rated value. At this moment, the automatic
    synchronizer will put the generator online by closing the generator
    After synchronizing the generators, the turbine jets will be
    closed (Fig.1. No flowoperation). At this stage, the machines will be
    rotating at the rated speed, and the synchronous generators will be
    operating as no-load motors. The power consumption in this
    operation mode will be very low (circa 2% of rated power), with
    only the power needed to compensate for the mechanical, iron and
    ventilation losses. If reactive power generation is required, this
    power consumption will be larger.
    In this situation, the Pelton units are ready to inject power in
    minimum time because they are already synchronized. If the power
    system requires a fast power injection, the turbine controller
    should be switched to power frequency regulation mode (Fig. 1.
    Frequency Regulation). In this mode of operation, power ramps are
    limited by the maximum allowable rate of flow change in the


    • guber says:

      Which makes the second failure only understandable with a cabel failure, because with the wind speed at that time they had 11,5 MW Wind +11,3 MW Hydro +2 MW Diesel available within seconds.
      Also at the first failure there were around 4MW Wind +11,3MW Hydro + 2 MW Diesel immediately available, so the grid should have been extremely stable – beside severe short circuits in the main power lines.

      • Kees van der Pool says:

        You may be right. However, the wind seems to have something to do with it too:
        2/18, wind=0 from 05:00 through 08:10 and 10:40 through 11.30
        2/19 wind=0 from 03:00 through 04:40
        These lacunae show up at the ‘disturbances’. To me, that sounds like the windpark was shut down for whatever reason. If this reason was a failing cable, the repair was not done very well and I hate to believe that.

  28. Rainer says:

    Solo Wind since 2015.02.20 1:10

  29. Rainer says:

    Wind standalone with water uphill since 2016.02.1:10

  30. Rainer says:

    Today observation:
    07:45 upper water storage not filled up but up to the former seen marks of max. water level.
    08:10 lower water storage nearly empty, water coming out of turbine house. Could hear turbines working. Forgot to listen to pumps in the pump house.
    08:40 upper storage: water level no visible change to 07:45
    09:10 lower storage: water level like 8:10. Sound: Pumps and turbines working.
    10:45 upper storage water level no visible change to 07:45
    11:10 lower storage water level like 8:10. Sound: Pumps and turbines working.
    Sound of pumps first quit same like 9:10, then a little less sound, after checking in front of the pump house same sound like 9:10.
    We should install a sound meter………..
    Pictures of water storages will be send to Roger and Hubert.

    • Kees van der Pool says:

      Excellent – very interesting. Running the turbines ‘wet’ keeps the minimum lower storage level under control and supplies instant spinning reserve as a bonus. Thanks.

  31. After three grid crashes in the last few days GdV has just completed another 24 hours of diesel-free operation. Never a dull moment.

  32. Euan Mearns says:

    11:00 am Sunday and the wind has died completely and they’re using the hydro – 56% of 5 MW demand. They now seem to be using the plant the way it was intended and this will make it more interesting to follow.

    According to Roger’s first post on El Hierro, the hydro can store about 50 MWh. Currently producing close to 3 MW it would run dry in 17 hours. The calm weather looks set to continue until mid-week. So I’ll be betting that they are back to 75% diesel by Monday morning – which is how the system was designed after all.

    • sod says:

      “11:00 am Sunday and the wind has died completely and they’re using the hydro – 56% of 5 MW demand. They now seem to be using the plant the way it was intended and this will make it more interesting to follow.”

      The plant is not used the way it was intended. The 11.5 MW wind is capped at 7 MW so far. That is pure insanity.

      One look at the output also shows, that they are doing just totally insane things. They should be compensating changes with water power and keep the diesel running at the best output level.


      those diagrams show that they are not making sense.

      ” So I’ll be betting that they are back to 75% diesel by Monday morning – which is how the system was designed after all.”

      The system was never designed for 75% diesel (why do you think it was and based on what evidence?)

      As Hubert has shown in this post, renewables made up 49.5 % in the first 2 months already.


      And this is with wind being capped AND the majority of water being used for irrigation instead of electricity production.

      Wind alone on this island should have a capacity factor of 40-50% and that for should cover demand (with water and ignoring the longer wind pauses) to 80- 90%.

      • Euan Mearns says:

        21:40 Sunday and they’re running 79% diesel.

        • sod says:

          Just look at the data from the last 2 days:

          They have been pumping up about 2MW for one day all of 20/02/2016


          Now they have barely used one MW all day long today.


          Running the diesel, even today, is by choice and not by necessity.

          And you simply keep ignoring that wind is capped at 7MW, which is a plain out joke.

          • Kees van der Pool says:

            Hi Sod,

            Thank you for your nuanced opinion about the operation of GdV.

            I trust you have taken into consideration that the major storm of last week was from the north and was gusting violently. Witness El Diario, Canarias Ahora: “la racha maxima del dia en Espana corresponde a la localidad de Sabinosa, en la costa noroeste de El Hierro, donde a las 15.00 horas (hora canaria) se registraron vientos de 114 kilometros por hora”. This could not have been fun for the windmills on the south east side of the island and standing in the eddy currents in the lee of a volcano at an altitude of 350 m or so. Maybe barreling along at rated power would not have been such a good idea under these circumstances. These mills are expensive and repair probably not trivial.
            Also, I hope you considered that after taking care of pumping and the grid requirements, excess power cannot be absorbed anywhere else and has to be curtailed.
            Of course, there is still the weather forecast playing into decisions, margins must be taken, the the top reservoir is twice as large as the bottom one etc., etc . . . .


  33. guber says:

    Well, with 4-7 MW demand and 6 MW possibility for pumping, the capping at 7 MW is not neccesary from the demand side. And how many broken E70 do you know, Kees? There are enough similar situations where the E70 are installed in similar environments without significant bigger maintenance needsas far as I have informations.
    For me it just looks like as if somebody with little experience and being very conservative or not liking the system is doing tests.

  34. Roger Andrews says:

    Sod and Guber:

    Please take the time to look into the constraints imposed by the grossly undersized storage reservoirs. Maybe then you can contribute some productive comments.

    • sod says:

      “Please take the time to look into the constraints imposed by the grossly undersized storage reservoirs. ”

      I do not think that the reservoirs are the limiting point of this system. If you think it is, please explain!
      The upper reservoir can basically run the system for close to a full day alone.

      as we see in the most recent data, they keep pumping up more water, than they use the next day.



      The worst failure is the cap of wind at 7 MW in a system that has peak demand of 7 MW.

      The second failure is the use of the majority of water for other stuff. You can not blame this system for not providing enough electricity, when they are simply using the water for something else.

      As the data clearly shows, those 40 hours could have been easily expanded to a full week of 100% renewables. It is by choice, that this is not done.

      The water reservoirs are a limit only, if you try to go to 100% renewables 100% of the time. But this is obviously not the idea behind the El Hierro system!

  35. guber says:

    @Roger – 150.000m² and 700m height means 1.03E12J or 286 MWh mechanical stored energy. At times when pumping goes on the next days, there is obvously enough to pump. So there is the possiblity to increase pumpung up to 6 MW, and later on use the water for balancing and switch of the Diesel.
    Obviously by some cause somebody is reluctant to do so, and limits the wind power output at 7MW, without obvious cause. (MAybe there is one which we don’t know, e.g. a too small cable etc.)
    I like your diagramms, but I do not always follow your conclusions. The reservoirs are big enough for roughly storing 40hours of electricity demand, and the upper reservoir can store a surpus of water for Irrigation in parallel. 40 Hours is too small for 100% wind power supply on a island with only local production. This is correct, it would fit better with a additional solar power generation.
    But this does not explain the curtailing at 7MW. Please explain why you think a curtailing at 7 MW is neccesary.

    • Rainer says:

      I agree with Sod and guber,
      the technology of GDV is not the problem. Of course not for really 100% all year.
      The Managemant of GDV i see as the main problem.

  36. Rainer says:

    2016.02.28 00:20 solo wind

  37. Sod: The REE graphs you are linking to are misleading. You have to plot diesel, wind and hydro separately from the REE tabulations to get a realistic idea of what is going on. I suggest you give it a shot.

    • sod says:

      I am not sure if i understand what you mean.

      The graphs can be changed by clicking on the green, blue and grey button on the left, removing and adding those electricity sources. By doing this (mostly i simply remove hydro), i get graphs that look exactly like yours at the top of this post.

      We now have another 18 hours without diesel again. the system is basically stable on wind alone (two really tiny dips into hydro at 7:40 and 13:30).

      At the same time, these plots completely contradict the claims being made about wind power, leading to extreme ramping up and down of fossile fuel plants. Diesel use is extremely stable most of the time.

      • primavera says:

        To see the real generationdata of REE:
        It do not have colors, but data to study

        Maybee somebody more competent can study the technology of hydrological short cut. I am really surprised how much energy is spoiled with this technology GDV use most of the time right now.
        Here i proudly present some data of the new frequency meter here in El Hiero:

        I selected a sequency where i suppose you can see the fight to hold the Grid constant.
        Think so much delta f is vary rare seen a big grid…..

        mHz Time
        -27.43 @UTC 18:44:48.576
        -23.13 @UTC 18:44:49.076
        -30.93 @UTC 18:44:49.576
        -61.13 @UTC 18:44:50.077
        -97.45 @UTC 18:44:50.577
        -130.70 @UTC 18:44:51.079
        -156.83 @UTC 18:44:51.580
        -180.42 @UTC 18:44:52.082
        -204.42 @UTC 18:44:52.584
        -205.51 @UTC 18:44:53.086
        -194.75 @UTC 18:44:53.588
        -195.59 @UTC 18:44:54.090
        -182.56 @UTC 18:44:54.592
        -166.08 @UTC 18:44:55.094
        -164.44 @UTC 18:44:55.595
        -102.24 @UTC 18:44:56.096
        -69.85 @UTC 18:44:56.597
        -44.44 @UTC 18:44:57.097
        -20.10 @UTC 18:44:57.597
        PS: still fiǵhthing putting this data online for everybody…….

  38. Rainer says:

    2016.02.29 11:00 Diesel started

    • sod says:

      So another 35 Hours basically running on wind alone.

      Has anyone taken a look at those dips during and after the test?

      They look artificial to me.

      (1:40 , 15:00 , 21:00)

      What are they testing? Was there some problem with one wind turbine?

      When will they finally go for a mix of wind and water, trying to go a longer time, for once?

      • Kees van der Pool says:

        “Wind flow over mountainous terrain may be relatively turbulent and may be outside the range for which commercial turbines have been designed,” said Manwell. “That could accelerate fatigue damage and premature failure and the need for replacement of some components. Thus reputable manufacturers may be loathe to sell turbines for such locations. There may also be problems getting financing or insurance.”

        It’s most cost effective to set up wind farms in plains regions and flat mesas, says David Minster, the Manager of the Wind Energy Technologies department at Sandia National Labs in Albuquerque. That is, places that have known winds and predictable directions. Although most wind turbines are designed to be able to change direction to face the wind, Minster says in mountainous areas they would have to change direction frequently.


        • Roger Andrews says:

          The GdV wind farm is oriented east-west along a ridgetop and production is good when the wind is from the north. But when it comes from the east or west production drops way off. You can see this in the El Hierro airport wind records.

          • Kees van der Pool says:

            Its not so much the production itself but the turbulence, shown in the graphs, that are violent. The very sharp light green multi-megawatt spikes downwards are clearly visible as is the consequent turning off of the pumps and sometimes kicking in of the hydro.

            There is a slew of papers available on the ‘3D’ wind component effect on windmills (mostly behind paywalls). I think the systematic curtailment at 7MW @ GdV has a lot to do with the fact that the mills are positioned on a ridge, 350m high, on a tiny island in the Atlantic, not in a steady breeze on the very flat Northern German coast. ‘Wear and tear’ comes to mind and if Enercon is responsible for the maintenance, they may have de-rated from the nominal rating at 15 m/s windspeed to allow for the ‘3D’ wind component. It cannot be fun to haul huge wings up the mountainside together with the appropriate crane. BTW, I don’t know if they are at 350m but they seem to be halfway between the top reservoir @ 700m and the harbor – maybe somebody with Google Earth Professional can confirm the altitude as the Google freebie has it greyed out.

          • sod says:

            wind came from north east or NNE.


            This is not the only wind farm ever being build or run. Wind farms on hills do just fine and with all the sea around, this one should do rather perfect.

            There is simply no reason for curtailment to 7MW in a 11.5MW system with pumped storage abilities and curtailment at even lower numbers makes even less sense.

            Today the output was extremely stable at 5MW, without any real reason. looks like curtailment to me.

          • Kees van der Pool says:

            ‘Reply’ button to sod disappeared but this would make interesting reading for him:

            Resolving Difficult Issues Of Wind Power Micrositing In Complex (2004):

            “Micrositing of wind turbines in complex terrain is tricky game, and many of the current siting methods and tools, while useful and improving, remain inadequate to the task in extreme terrain. As a consequence, there are numerous wind turbine installations that are buffeted by damaging turbulence or are faced with suboptimal wind energy performance”


  39. sod says:

    Looks like wind is curtailed at 5MW now. Are they moving backward or forward with this project?


  40. Rainer says:

    2016.03.02 19:00 wind 0m/s
    This wind calms did know Columbus too……
    El Hierro is located at the north rim of the north-western trade winds……
    Local people say the trade winds are not as regularly as they liked to be……

Comments are closed.